NXP backend: Test max_pool2d with new Neutron flow.

[MartinPavella]

Summary

Reflect the requirements of the new Neutron MLIR flow for the max_pool2d operator in NXP backend.

Test plan

Unit tests provided.

cc @robert-kalmar @JakeStevens @digantdesai

[pytorch-bot]

🔗 Helpful Links

🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/executorch/19272

• 📄 Preview Python docs built from this PR

Note: Links to docs will display an error until the docs builds have been completed.

❗ 1 Active SEVs

There are 1 currently active SEVs. If your PR is affected, please view them below:

• Ubuntu services are down

❌ 1 New Failure, 2 Unrelated Failures

As of commit a3f359a with merge base 48a8d58 ():

NEW FAILURE - The following job has failed:

• pull / unittest-nxp-neutron / linux-job (gh)
  RuntimeError: Command docker exec -t 223fab041b1ce288f497803d80eb28d9771b397bd671050ff767bf178c240e95 /exec failed with exit code 1

BROKEN TRUNK - The following jobs failed but were present on the merge base:

👉 Rebase onto the `viable/strict` branch to avoid these failures

• pull / unittest / macos / macos-job (gh) (trunk failure)
  ##[error]The operation was canceled.
• pull / unittest-editable / macos / macos-job (gh) (trunk failure)
  ##[error]The operation was canceled.

This comment was automatically generated by Dr. CI and updates every 15 minutes.

[MartinPavella]

@novak-vaclav I am not able to add you as a reviewer, but feel free to have a look.

[MartinPavella]

The failing Neutron job is caused by temporary outage of the University of Toronto website which provides the Cifar10 dataset for NXP tests: https://www.cs.toronto.edu/~kriz/cifar.html
It should be fixed later today.